Thinking... A Customer Information System

- Gerry Gaffney. July 24, 2002.

Download 'printer-friendly' PDF version (File size: 54 KB)

Introduction

In Melbourne, Australia, many railway stations have a device labeled as providing 'Customer Information'.

The device, situated on a railway platform, consists of two buttons and a speaker, mounted on an instructional panel.

One button is for 'next train information', one is for 'emergency assistance'. When the green button is pressed, information is provided about train departures. The information is delivered by a recorded voice via a phone line from a central location. The information is 'assembled' from pre-existing snippets.

The following is a typical message. The square brackets indicate the assembled snippets.

[Trains departing] [Toorak] [Platforms 1 and 2] [are] [the 9:15 Flinders St] [stopping all stations to Flinders Street] [via the City loop] [departing] [platform 1] [in 12 minutes]

'[the 9:20 Frankston] [stopping all stations to Frankston] [departing] [platform 2] [ in 17 minutes] [...]

'[Remember to validate your ticket before traveling].'

Message duration is typically at least 30 seconds.

Benefit

The benefit of this system is that it provides real-time information. Prior to installation of these devices, only the information in the published timetables was available. Late-running or cancelled trains are relatively common, so this information was often inadequate.

The system provides some degree of support on stations that no longer have staff members present to assist and provide information.

Suspected Usability Issues

A very brief 'heuristic evaluation' of the device reveals many issues. Some of these are as follows.

Non-redundant use of color

The instructional text refers to the 'green button' (for information) and the 'red button' (for emergency assistance). However, the buttons are not labeled. Since a relatively high percentage of males suffer from red-green color deficiency, it is likely that many people may not be able to distinguish between the buttons. This is potentially dangerous in an emergency situation. One might infer that since the green button is referred to first, it would be on the left. It is, however, on the right.

Lack of support for casual users

While travelers who are familiar with the railway station may know which platform to use, visitors or tourists must go to each platform in turn, and query each machine in turn, in order to find out when and where to catch the appropriate train. It may be argued that this information should available elsewhere - in some cases, however, it is not.

Misleading labeling

The device purports to cater for 'General Enquiries'. However, there is no apparent mechanism for handling general enquiries. (In fact, using the red 'emergency' button contacts a human operator who may provide some assistance.)

Provision of irrelevant information

Where platforms are adjacent, the device gives information about trains from both platforms, even though in many cases the destinations are different; hence much of the information is typically irrelevant to the user.

A more detailed analysis would reveal several other issues. For example, the system appears to be a poor channel for delivering admonitory information such as the request to validate tickets before traveling, since users may have stopped listening by the time that information is delivered.

Observed Usability Issues

Observing the device in use on a casual basis over several weeks confirms the issues already identified, and reveals some others.

Color

Conversations with apparently puzzled male users on separate occasions reveals that the green and red buttons are indistinguishable to some users.

Poor support for multiple users

Because the announcements are not loud, only a relatively small number of people can hear them. This leads to repeated usage of the device.

Susceptibility to noise

The announcements become inaudible if a train is in motion at a nearby platform.

Poor availability

The device relies on a telephone connection. The lines are frequently busy, in which case the user simply hears a busy (engaged) tone until the connection times out. This issue is exacerbated by the poor support for multiple users and susceptibility to noise, both of which lead to frequent usage attempts, particularly during peak hours.

Addressing the Issues

Several of the issues identified during our rather cursory review lend themselves to 'quick fixes'. For example, labeling the buttons would resolve the red/green confusion issue readily. Providing additional telephone capacity could reduce the incidence of 'busy' signals.

Some of the issues may be the result of technology constraints. For example, the decision to use voice for delivery of the timetable information (as opposed to using visual displays) may have been made as a result of availability of appropriate servers. If this is the case, the issues related to the medium cannot be readily addressed. For example, the volume of the message could be increased, but only within a fairly narrow range.

Systemic issues

Some of the issues, however, are systemic, and cannot be addressed at the device level.

Examining these systemic issues quickly leads us beyond the device itself, and into the way-finding and navigational support within the rail system as a whole. There appear to be two underlying issues:

• Lack of an integrated travel information system. The rail system has been divided up and privatized. Individual companies look after their own 'turf', and appear not to have an interest in integration. For example, maps may show only the rail lines run by a specific company; websites have different look-and-feel, and in-train notices and displays use different terminology.



• Reduction of staff numbers leads to an increased reliance on support systems.

What can we learn?

There are some key points we can gather from this exercise:

• The application of user-centered design techniques might have produced an alternative design. For example, the question 'What platform should I go to for a train to the City?' is not effectively answered by these devices. Simple techniques such as the use of scenarios and personas would have been appropriate.



• Technology solutions may only be 'band-aid' if underlying or systemic issues are not addressed. The devices offer previously-unavailable information but are not integrated within an overall strategy. Such solutions are doomed to be piecemeal and of limited efficacy.



• Some improvements are possible by focusing on the user interface of the individual device or system.



• Effective design of usable systems is dependent on first winning the appropriate level of political and organizational support.

About this Series

The 'Thinking...' series is not intended to simply complain. Real-world constraints often force designers to make choices that are not optimal from a usability perspective, and it is all too easy for an outsider to be harshly critical. The series is an attempt to analyze and learn by examining existing designs. Contact Gerry Gaffney with comments or suggestions.

Download a 'printer-friendly' PDF version of this sheet (File size: 54 KB)

Download Adobe's free Acrobat Reader (TM) for PDF files